School Gender Composition and Academic Performance: Evidence from Transition from Single-Sex to Coeducational Schools

Abstract

We examine how the gender composition of students within schools affects their academic performance. For causal identification, we exploit within-school variation in the gender composition because of policy-driven transitions from single-sex to coeducational schools. In Seoul, South Korea, several high schools were converted from single-sex to coeducational schools between 1998 and 2003, by the city superintendents Coeducational School Expansion Policy. We find that boys test score dropped when their schools began to admit girls based on administrative test score data on standardized college entrance examination. However, the negative effect disappeared when the school transition was complete. We find no effect on girls who were admitted to previously boys-only schools.